Sustainable production of dimethyl adipate by non-heme iron(iii) catalysed oxidative cleavage of catechol

Article abstract of DOI:10.1039/c4cy01562b

Adipic acid and its esters are important bulk chemicals whose principal use is in the production of the nylon-6,6 polymer. There is considerable interest in finding novel green routes from sustainable feedstocks towards these important intermediates. Herein, we describe the catalytic oxidative cleavage of catechol to muconic acids using a catalyst prepared in situ from iron(iii) nitrate, tris(2-pyridylmethyl)amine and ammonium acetate. An investigation of catalyst loading, temperature and oxygen pressure, allowed a turnover frequency of 120 h^-1 to be obtained. The subsequent hydrogenation and transesterification of the obtained muconic acid products were shown to proceed well over commercially available supported catalysts. After vacuum distillation, dimethyl adipate could be isolated in 62% yield from catechol, thus demonstrating a green and sustainable route to this important bulk chemical.

Full text of DOI:10.1039/c4cy01562b

Catalysis
Science &
Technology
View Article Online
View Journal
PAPER
Sustainable production of dimethyl adipate
by non-heme ironIJIII) catalysed oxidative
cleavage of catechol†
Cite this: DOI: 10.1039/c4cy01562b
Robin Jastrzebski, Emily J. van den Berg, Bert M. Weckhuysen
*
and Pieter C. A. Bruijnincx
Adipic acid and its esters are important bulk chemicals whose principal use is in the production of the
nylon-6,6 polymer. There is considerable interest in finding novel green routes from sustainable feedstocks
towards these important intermediates. Herein, we describe the catalytic oxidative cleavage of catechol to
muconic acids using a catalyst prepared in situ from ironIJIII) nitrate, trisIJ2-pyridylmethyl)amine and
ammonium acetate. An investigation of catalyst loading, temperature and oxygen pressure, allowed a
turnover frequency of 120 h−1 to be obtained. The subsequent hydrogenation and transesterification of the
obtained muconic acid products were shown to proceed well over commercially available supported
catalysts. After vacuum distillation, dimethyl adipate could be isolated in 62% yield from catechol, thus
demonstrating a green and sustainable route to this important bulk chemical.
Received 27th November 2014,
Accepted 15th December 2014
DOI: 10.1039/c4cy01562b
www.rsc.org/catalysis
Van de Vyver and Román-Leshkov recently reviewed the
development of novel catalytic processes for adipic acid
1. Introduction
Adipic acid and its diesters are the most important carboxylic
diacids produced in the chemical industry, with a production
of approximately 2.8 million tons per year. The most important
application is the production of the nylon-6,6 polyamide,
although other uses include synthesis of polyesters and
polyurethanes, plasticisers in polyvinyl chloride and polyvinyl
butyral and as additives in numerous other applications.1
The current industrial production of adipic acid (Fig. 1)
starts from oil-derived benzene, which is first hydrogenated
to cyclohexane. Cyclohexane is typically oxidized by air to
KA-oil, which is a mixture of cyclohexanol and cyclohexanone,
over a cobalt catalyst.2 As selectivity is inversely dependent on
cyclohexane conversion, this reaction is typically run at low
(4–8%) conversion and requires continuous distillation and
recycling of cyclohexane.3 Subsequent oxidation of the KA oil
with nitric acid affords adipic acid, as well as a stoichiometric
amount of nitrous oxide, a strong greenhouse gas.4,5 There is
considerable interest in novel processes for the production of
adipic acid, both to abate waste emissions and to expand the
feedstock to sustainable alternatives for benzene.
production.6 The furthest developed alternative to the current
industrial production is the oxidation of cyclohexene with
30% hydrogen peroxide with a catalyst based on soluble tung-
states, which was first reported by Sato et al.7 Several routes
starting from renewable substrates have also been reported, for
example from 5-hydroxymethylfurfural,8 γ-valerolactone9 and
D-glucose.10 In the latter, biochemical, route E. coli strains were
engineered to convert glucose into cis,cis-muconic acid, which
could in turn be readily converted into adipic acid by catalytic
hydrogenation. Lignin could also serve as a renewable resource
for adipic acid production, as this renewable material is
available at low-cost and in large quantities, but lignin-based
routes are still largely unexplored.11
With lignin valorization in mind, catechol provides an
interesting starting material. Indeed, guaiacols (i.e. catechol
methyl ethers) are often reported as major products in lignin
depolymerisation procedures11–14 and hydrolysis may then
afford the corresponding catechols.15–18 In addition, lignin
depolymerisation procedures that directly yield catechols have
recently been reported.19
The catechol-based route reported here (Fig. 1) is inspired
by Nature, with 1,2-catechol dioxygenase enzymes, for instance,
being able to convert catecholic substrates to cis,cis-muconic
acids. This enzyme belongs to the class of intradiol catechol
dioxygenases, which selectively cleave the carbon–carbon
bond of 1,2-dihydroxybenzenes between the two hydroxyl
functionalities, using molecular oxygen.20–23 The enzyme
Inorganic Chemistry and Catalysis, Debye Institute for Nanomaterials Science,
Utrecht University, Universiteitsweg 99, 3584 CG Utrecht, The Netherlands.
E-mail: p.c.a.bruijnincx@uu.nl
† Electronic supplementary information (ESI) available: 1H NMR spectra,
analytical data for dimethyl adipate, absolute energies and Cartesian coordinates
for calculated structures. See DOI: 10.1039/c4cy01562b
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
Fig. 1 The current industrial production of adipic acid starts from benzene, which is hydrogenated to cyclohexane. Aerobic oxidation using a
cobalt catalyst affords KA oil, which is oxidised by nitiric acid to adipic acid. Our novel route starts from pyrocatechol, which is oxidized to the
corresponding muconic acid by the iron catalyst. Subsequent hydrogenation and transesterification affords dimethyl adipate.
active site consists of an ironIJIII) centre in a non-heme
2. Results and discussion
coordination environment, where the substrate binds to iron
as a dianion.24
The dioxygenation of pyrocatechol (1) using an in situ
prepared ironIJIII) trisIJ2-pyridymethyl)amine (TPA) complex
(Fig. 1) was investigated in a stainless steel batch autoclave
equipped with reflux condenser to allow reactions to be
run at elevated pressures and under continuous gas flow. In
our previous communication29 we already showed that the
highest activities are obtained with ironIJIII) salts with non-
coordinating anions, using mild bases; we therefore used
readily available ironIJIII) nitrate nonahydrate as the iron
source here. Variation of base concentration showed that the
highest activity was obtained with two equivalents of base
relative to the catalyst. In order to prevent undesirable side
reactions, the (very) weak base ammonium acetate was
employed. The TPA ligand is commercially available, but can
also be conveniently prepared by a literature procedure.30 In
all oxidation chemistry, including industrially practiced pro-
cesses, such as the cyclohexane oxidation step of the current
adipic acid process, one should be aware of the potential fire
and explosion hazards that result from the use oxygen and
flammable organic compounds and solvents under pressure
and elevated temperature. Therefore, all experiments were
run using a 1 : 3 mixture of synthetic air and nitrogen, for a
final oxygen concentration of 5 vol%. This is well below the
limiting oxygen concentration for a methanol/oxygen system
and thus outside the flammability range.31
A typical reaction proceeded for 4 h at 50 °C and 10 bar
total pressure, using 0.5 mol% catalyst. Products were iso-
lated from the reaction mixture by acidification with aqueous
HCl to dissociate the iron complexes and subsequent extrac-
tion with diethyl ether. In all reactions, a maximum of six
products was observed, which all derive from the selective
intradiol dioxygenation of catechol (Fig. 2). Products that
derive from ring cleavage with a different regioselectivity
(i.e. extradiol dioxygenation) were not observed. Thus, due to
the convergent nature of the subsequent hydrogenation and
esterification steps, all six products are desired, as they will
all be converted to dimethyl adipate.
Biomimetic synthetic ironIJIII) complexes bearing nitrogen
and oxygen donor ligands have been developed as functional
models for these intradiol cleaving dioxygenases and are
able to perform this reaction.25–27 The best performing com-
plex in the stoichiometric intradiol dioxygenation of the acti-
vated model compound 3,5-di-tert-butylcatechol is the system
based on trisIJ2-pyridylmethyl)amine (TPA) reported by
Que et al.28 Recently, we communicated the first example of
catalytic intradiol dioxygenation reaction of pyrocatechol
using an ironIJIII) TPA complex as catalyst, albeit still with
somewhat modest turnover numbers.29
Herein, we report our further investigation into this reac-
tion, which enabled a 35-fold improvement in the turnover
frequency of the catalyst. The reaction was shown to proceed
at catalyst loadings as low as 0.1 mol%, while an increase in
reaction pressure and temperature significantly increased
activity with retention of the excellent selectivity towards the
desired oxidative cleavage products. Based on DFT calcula-
tions we show that isomerisation of the products takes place
as a secondary, uncatalysed, reaction through a lactonic tran-
sition state. Furthermore, we investigated the sequential con-
vergent hydrogenation and transesterification of the obtained
product mixtures to dimethyl adipate, using compatible reac-
tion conditions and recoverable catalysts, minimising the
amount of work-up and isolation required in the reaction
sequence.
Compared to the current industrial route, our new route is
highly atom-efficient: water, rather than nitrous oxide, is the
only stoichiometric by-product and only two equivalents of
hydrogen (compared to three in the industrial process) are
necessary to obtain the final adipates. In contrast to the high
temperatures and pressures required for benzene hydrogena-
tion in the conventional route, dearomatisation here pro-
ceeds oxidatively and as a result all reaction steps can be run
at mild temperatures and pressures. Furthermore, our pro-
cess uses largely non-toxic reagents and catalysts. The route
thus provides a sustainable alternative to dimethyl adipate
production.
The cyclic muconic anhydride, which is the initial product
of the intradiol dioxygenation reaction, is not directly observed.25,28
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
the cis,cis isomer. This appears to be a reasonable activation
barrier considering the relatively low cis,trans isomer yields that
are observed. In the dioxygenation of substituted catechols,
lactones have previously been observed as products,26,34,35
although we find none under our conditions. The absence of a
lactonic intermediate also explains why no lactone products
are observed, as these might otherwise be formed simply by
protonation of such an intermediate.
Fig. 2 Products observed in the catalytic dioxygenation of catechol are
derivatives of cis,cis-muconic acid (2) and cis,trans-muconic acid (3). The
products include free acids (a), methyl esters (b) and catechol esters (c).
Under standard conditions, the six products and
unreacted catechol accounted for 99% of the original catechol.
Under more severe reaction conditions, we were unable to
completely close the mole balance (as moles of C6-units found,
against the original intake of C6 units) and ca. 5–15% of
the original catechol intake was unaccounted for. To ensure
all muconate products were completely extracted from the
reaction mixture, the aqueous fractions were analysed by
UV-vis measurements after extraction. The muconate species
have very intense absorption bands between 251 nm and
264 nm,36 but no strong absorbances other than those from
the TPA ligand could be observed (Fig. S7†). Having now
excluded muconate losses due to the work-up procedure, we
would attribute the reduced mole balance to noncatalytic
oxidation to 1,2-benzoquinone, as we previously hypothesised.
This highly unstable compound is known to undergo conden-
sation reactions with catechol itself, forming poorly-defined,
water-soluble oligomers, which are as a result not extracted
and seen in our work-up procedure.37–39
Rapid nucleophillic attack by water, methanol or catechol
affords, respectively, cis,cis-muconic acid (2a), its monomethyl
ester (2b) and its catechol ester (2c). The major product
observed in all cases is 2b. At 96% conversion, a yield of 63% is
observed, while yields of 2a and 2c are only 5% and 1% respec-
tively. Isomerisation to the corresponding cis,trans-isomers
(3a–c) is also observed, with a typical 2 : 3 ratio of 7 : 1.
The isomerisation of muconic acids has previously been
proposed to proceed by intramolecular attack of the carboxylate
moiety on the 1,3-diene to form a lactonic intermediate.32,33
Given the rather mild conditions at which the reaction is
performed, we used density functional theory (Amsterdam
Density Functional, BP86 functional, TZP basis set) to check if
such a mechanism would be feasible under our reaction condi-
tions. Geometry optimisation of the monomethyl muconate
anion indeed gave minima for the cis,cis and cis,trans isomers.
As expected, the cis,trans isomer is the thermodynamic end
product having a free energy that was slightly lower than the
cis,cis isomer by 2.4 kcal mol−1. Interestingly, however, no
stable lactonic intermediate could be found, but increased
lactonic character of the muconic acid did lengthen the
carbon–carbon double bond. Consequently, we were able to
optimise a lactonic transition state for the isomerisation
(Fig. 3) with a free energy of 23.7 kcal mol−1 compared to
2.1 Catalyst
As we have previously performed the intradiol dioxygenation
at a rather high catalyst loading of 5 mol%, we were inter-
ested to see if the catalyst retained its activity at lower con-
centrations. Thus, the dioxygenation reaction was performed
for 4 h of reaction time using loadings of 0.1–5 mol% catalyst
at 50 °C and 10 bar total pressure (Table 1). At high catalyst
loadings, the reaction proceeded with full conversion, while
at lower catalyst loadings the turnover frequency increased up
to 27 h−1 at 0.1 mol%. As the catechol intradiol dioxygenation
proceeds approximately 0th order in substrate,29 the increase
in turnover frequency presumably is the result of the 0th
Table 1 Results of the catalytic dioxygenation of catechol using different
amounts of catalysta
Cat.b
(mol%)
Conv.c
(%)
TOFd
MBe
(%)
Yield 2 + 3 (%)
(h−1
)
a
b
c
1
2
3
4
5
6
0.1
0.5
0.5f
1.0
2.0
5.0
12
40
43
90
100
100
27
20
18
19
10
8
99
99
93
91
92
86
1
4
3
6
4
3
10
33
28
61
71
64
<1
2
4
7
10
9
a
Conditions: 27.2 mmol catechol, catalyst (1 eq. FeIJNO3)3, 1 eq. TPA,
Fig. 3 Transition state (top, selected bond lengths indicated in Å) and
mechanism (bottom) for double bond isomerisation in monomethyl
muconate as revealed by density functional theory.
2 eq. NH4OAc), 300 mL MeOH, 50 °C, 10 bar (5% O2), 800 sccm gas
b
c
d
flow, 4 h. Catalyst loading. Conversion. Turnover frequency.
e
f
Mole balance. Using ironIJIII) oxalate and no NH4OAc.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
order approximation being more valid at low conversions,
rather than a true activity increase. The product distributions
are largely unaffected by the concentration of the catalyst.
To further simplify the system and reduce the environ-
mental footprint by omitting the externally added base, we
were also interested in testing the activity of an ironIJIII) salt
with weakly basic counterions. As the first turnover produces
muconate anions which then in turn act as base to continue
the catalytic cycle, the basic counter ions merely are needed
to initiate the catalytic cycle and a pKa ~3 of the conjugate
acid should be sufficient. Thus, when performing the reac-
tion using ironIJIII) oxalate (Table 1, entry 3) without any
added base, an activity that was only slightly lower than with
2 eq. of NH4OAc was obtained. The oxalate anions are also
sufficiently basic to enable formation of the first catecholato
complexes that initiate the catalytic cycle and also do not
appear to significantly displace the catecholato moieties from
the ironIJIII) centre.
2.3 Reaction temperature
The influence of temperature (30–80 °C) on the reaction was
investigated at 10 bar total pressure (5% O2) and 0.5 mol%
catalyst (Table 3). As observed previously for the influence
of catalyst loading and pressure, no significant changes in
product selectivity were observed. Activities clearly increase
with increasing temperature, although the mole balances do
decrease somewhat with increasing conversion.
At first glance, the observed activities appear to increase
approximately linearly with temperature, which is attributed
to the rather narrow absolute temperature interval which was
measured. To better interpret the dependence of the rate as a
function of temperature, the data was plotted in the form of
the Eyring–Polanyi equation (eqn (1), Fig. 4):
k
H #
R
1
kb S#
(1)
ln  
 ln

T
T
h
R
From the slope and intercept, the standard enthalpy and
entropy of activation were determined to be 23.8 kJ mol−1
2.2 Reaction pressure
In order to increase the activity of the system, the influence
of the total gas pressure (5–25 bar at 5 vol% oxygen) was also
investigated, using 0.5 mol% catalyst at 50 °C (Table 2). An
approximately linear increase of activity with pressure was
observed, suggesting a first-order process in oxygen. This is
consistent with the proposed mechanism for catechol intra-
diol dioxygenation that involves attack of oxygen on the
active iron catecholato complex40,41 and with our recent
mechanistic work on this system.42 To exclude the influence
of mass-transfer of oxygen into the liquid phase, which is
also a first-order process, a control experiment with lower
stirring rate was performed. The similar activity that was
observed, together with the intrinsic low oxygen conversion
(ca. 3%) in our experimental conditions, make it seem rea-
sonable to assume that we are not working under mass-
transfer limited conditions and that the pressure-dependent
data represent the actual, expected order in oxygen. As with
the catalyst loading, no significant changes in product distri-
bution are observed with increasing pressure, which is also
expected, as 2 and 3 are products of secondary reactions.
Table 3 Results of the catalytic dioxygenation of catechol using different
temperaturesa
Tb
(°C)
Conv.c
(%)
TOFd
MBe
(%)
Yield 2 + 3 (%)
(h−1
)
a
b
c
1
2
3
4
5
6
7
8
30
40
50
60
70
75
80
80f
21
45
43
75
84
96
96
63
8
94
90
99
84
76
75
87
84
1
3
4
4
7
5
8
4
14
25
33
45
53
62
69
41
0
4
2
5
0
2
3
2
16
20
28
30
35
40
120
a
Conditions: 27.2 mmol catechol, 0.5 mol% catalyst (1 eq. FeIJNO3)3,
1 eq. TPA, 2 eq. NH4OAc), 300 mL MeOH, 10 bar (5% O2), 800 sccm
b
c
d
gas flow, 4 h. Temperature. Conversion. Turnover frequency.
e
f
Mole balance. 0.1 mol% catalyst, 25 bar pressure.
Table 2 Results of the catalytic dioxygenation of catechol using different
total pressures (5% O2).a
pb
(bar)
Conv.c
(%)
TOFd
MBe
(%)
Yield 2 + 3 (%)
(h−1
)
a
b
c
1
2
3
4
5
6
5
29
43
56
73
88
97
11
20
23
31
36
37
96
99
94
94
86
96
2
4
4
6
6
8
18
33
37
51
63
66
2
3
4
5
2
<1
10
10f
15
20
25
a
Conditions: 27.2 mmol catechol, 0.5 mol% catalyst (1 eq. FeIJNO3)3,
Fig. 4 Temperature dependence of the reaction rate of catechol
intradiol dioxygenation expressed as an Eyring–Polanyi plot and the
derived thermodynamic parameters.
1 eq. TPA, 2 eq. NH4OAc), 300 mL MeOH, 50 °C, 800 sccm gas flow,
b
c
d
e
4 h. Pressure. Conversion. Turnover frequency. Mole balance.
f
400 rpm stirring.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
and −172 J mol−1 K−1, respectively. The large negative entropy
of activation can be attributed to the binding of oxygen to the
active catalyst complex, which is also not very favourable in
terms of enthalpy and is thus an important limiting factor in
the reaction rate. This high ΔS# is also in good agreement
with our recent computational study on this system,42 where
we found an entropy of activation of −182 J mol−1 K−1. Never-
theless, increasing the temperature from 30 °C to 80 °C
afforded a five-fold increase in the reaction rate.
The transesterification of the hydrogenated acid mixture
with methanol was previously successfully performed using
p-toluenesulfonic acid as catalyst. Using a commercially
available heterogeneous solid acid catalyst (Amberlyst-15) in
refluxing methanol we were also able to (trans)esterify both
free acids and catechol esters, to obtain dimethyl adipate.
Purification by vacuum distillation (14 mbar, 80–105 °C) then
affords highly pure dimethyl adipate in 62% overall yield.
Analysis of the tarry distillation residue showed significant
quantities of dimethyl adipate to be still present, which we were
unable to recover given the rather small scale on which the
distillation was performed. Furthermore, the distillation resi-
due still contained the catechol liberated by transesterification
of the catechol esters, which thus in principle could be recycled
for another oxidation reaction.
Consequently, the main losses in the conversion of
catechol to dimethyl adipate using our three-step route seem
to occur by the radical oxidation of catechol during the first
oxidation step, while hydrogenation and transesterification
occur essentially quantitatively. Better recovery of the dimethyl
adipate during the final step is rather a technical problem,
which should be easily solved by better design of the distilla-
tion setup.
2.4 Improved conditions
Having investigated catalyst loading, pressure and temperature
separately, we were also interested in obtaining maximal activ-
ity in the system. Combining a low catalyst loading (0.1 mol%)
with high pressure (25 bar) and high temperature (80 °C), we
were able to attain a TOF of 120 h−1 and a turnover number of
480 (Table 3, entry 8), which is a 35-fold improvement over our
previously communicated system.
2.5 Hydrogenation and transesterification
For the investigation of the subsequent hydrogenation and
transesterification steps to obtain dimethyl adipate, product
mixtures were prepared from 3 g of catechol, using a standard
oxidative cleavage condition, i.e. 4 h reactions with 0.5 mol%
catalyst at 25 bar pressure and 80 °C to ensure full conversion
of catechol. After isolation, a mixture of products is obtained
as a solid in 85% yield (assuming an average molecular
weight of 156.14), with full conversion and 3b as the domi-
nant product. As the hydrogenation and transesterification
step converge all products to dimethyl adipate, no further
separation of the products is required nor performed.
3. Experimental
3.1 Materials
Catechol (99+%) and ironIJIII) nitrate nonahydrate (98+%)
were purchased from Acros. Ammonium acetate (HPLC grade)
was obtained from Fisher Scientific. IronIJIII) oxalate hexa-
hydrate, 1-dodecanol (98%), RANEY® nickel (RANEY® 4200),
palladium supported on carbon (5 wt% Pd) and Amberlyst-15
(dry, acid form) were obtained from Sigma Aldrich. Methanol
( p.a. grade) was obtained from Merck. Compressed air (4.0)
and nitrogen (5.0) were obtained from Linde Gas. All chemicals
were used as received. TrisIJ2-pyridylmethyl)amine was synthesised
according to a previously published procedure30 and crystallised
from boiling hexanes.
The hydrogenation reaction was investigated using
commercially available catalysts. RANEY® nickel is a cheap
and readily available hydrogenation catalyst, although often
elevated temperatures and hydrogen pressures are necessary to
obtain good activity. Nevertheless, we found that our mixtures
were hydrogenated over RANEY® nickel at very mild conditions
(16 h, 1 bar H2, room temperature), although some half-
hydrogenated muconic acid products (including the 3-ene
1
3.2 Physical measurements
isomer) were identified based on their H NMR signals. After
1
simple filtration to remove the catalyst, the H NMR signals of
NMR spectra were recorded on a Varian spectrometer operat-
ing at 400 MHz (1H) or 100 MHz (13C) in DMSO-d6 (unless
mentioned otherwise) and referenced against the signal of
the residual protio impurity of the solvent (2.50 ppm). GC-MS
measurements were performed on a Shimadzu GC-2010 using
a VF5-ms column, coupled to a Shimadzu GCMS-QP2010 mass
spectrometer.
the products were found to be extremely broad, suggesting the
presence of paramagnetic species, possibly as Ni2+ carboxylates,
in the sample. An additional acid wash step was thus required
to obtain a clean hydrogenated mixture, after which the adipates
were isolated in 57% yield.
A commercial palladium on carbon catalysts also readily
hydrogenates the muconic acid mixture under the same mild
conditions. Reflecting the higher activity of carbon-supported
palladium, only fully hydrogenated products were obtained
in near quantitative yield. As no half-hydrogenated products
are observed in this case, this method is preferred for the
hydrogenation. Nevertheless, in an industrial setting where
more severe conditions to reach full conversion would not be
limiting, the use of cheaper RANEY® nickel may be viable.
3.3 Catalytic studies
Catalytic oxidation reactions were carried out in a fed-batch
500 mL stainless steel autoclave reactor setup (SA182-F316,
rated up to 254 bar at 450 °C) by Autoclave Engineers. Gas
inlet was controlled by Brooks mass flow controllers and gas
outlet by a needle valve to maintain a constant pressure. The
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.

View Article Online
Paper
Catalysis Science & Technology
effluent gas was cooled by a reflux condenser in order to
minimise solvent losses.
dimethyl adipate (2.92 g, 17.2 mmol, 62%) as a pale yellow
1
1
oil (see Fig. S3–S5† for H, 13C and GC-MS analysis). H NMR
analysis of the tarry distillation residue (Fig. S6†) showed that
it still contained a significant quantity of dimethyl adipate.
In
a typical experiment, the autoclave was charged
with catechol (3.00 g, 27.2 mmol), aliquots of stock solutions
of ironIJIII) nitrate (0.136 mmol), trisIJ2-pyridylmethyl)amine
(0.136 mmol) and ammonium acetate (0.272 mmol),
1-dodecanol as internal standard (3.00 g, 16.1 mmol) and
methanol (300 mL). The autoclave was sealed, purged with
nitrogen and heated to a stable operating temperature. Oxy-
gen was introduced by mixing compressed air with nitrogen
(800 sccm gas flow, 5 vol% oxygen), marking the start of the
reaction. After 4 h, gas flow was switched to pure nitrogen
and the reactor was cooled to room temperature. A 1 mL
sample was partitioned over 5 mL 1 M HCl and 5 mL diethyl
ether. The organic layer was separated and the aqueous
phase extracted two more times with 5 mL diethyl ether.
The combined organic layers were dried over magnesium
sulphate, filtered and concentrated in vacuo. The residual oil
was dissolved in DMSO-d6 and analysed by NMR spectro-
scopy. Products were quantified by their previously identified
characteristic chemical shifts.29 Turnover frequencies were
determined by taking the total molar yield of intradiol cleav-
age and dividing by catalyst amount and reaction time. The k
values (for use in the Eyring–Polanyi equation) were obtained
by dividing the turnover frequency (in s−1) by the oxygen
concentration, which, based on Henry's law and the solubility
of oxygen in methanol, was assumed to be 0.5 mM at 10 bar
total pressure.43
3.4 Computational methods
Density functional theory (DFT) calculations were performed
with the Amsterdam Density Functional (ADF 2013.01)
package,44–46 using the BP86 functional47,48 and the triple
zeta Slater orbital basis set with polarisation functions (TZP)
that is included with ADF.49 Energies were corrected for sol-
vation using the conductor-like screening model (COSMO)50
and for dispersion interactions using Grimme's DFT-D3
dispersion correction51 with Becke–Johnson damping.52 All
structures were subjected to geometry optimizations without
any constraints followed by a full analytical frequency calcu-
lation. All reported structures were characterized by no (minima)
or one (transition states) imaginary frequencies. Energies
reported are Gibbs free energies at 298 K relative to the mono-
methyl cis,cis-muconate anion. Absolute energies for all struc-
tures are provided in the ESI.†
4. Conclusions
We have demonstrated the catalytic dioxygenation of catechol
to derivatives of cis,cis-muconic acid using 5% oxygen in
nitrogen, using a catalyst prepared in situ from ironIJIII)
nitrate, trisIJ2-pyridylmethyl)amine and ammonium acetate.
The reaction was shown to proceed at catalyst loadings as
low as 0.1 mol% and activity was shown to be approximately
linear with respect to oxygen. Investigating the temperature
dependence of activity, a high entropy of activation was found,
consistent with the need of oxygen to bind the active ironIJIII)
catecholate complex. Nevertheless, an increase in temperature
from 30 °C to 80 °C afforded a five-fold increase in reaction
rate and a turnover frequency of 120 h−1 could be obtained.
Subsequent hydrogenation and transesterfication of the
afforded muconic acid mixtures proceeded efficiently over
commercially available hydrogenation and solid acid cata-
lysts. After vacuum distillation, pure dimethyl adipate could
be recovered in 62% yield from the catechol starting material,
thus demonstrating a sustainable synthesis of this high vol-
ume nylon monomer.
Products for further hydrogenation and transesterification
were prepared as above (at 80 °C and 25 bar pressure), but
1-dodecanol was omitted from the reaction mixture. After
reaction, the methanol solution was concentrated to ca. 75 mL
and partitioned over 100 mL 1 M HCl and 100 mL diethyl
ether. The organic layer was separated and the aqueous phase
extracted two more times with 100 mL diethyl ether.
The combined organic layers were dried over magnesium
sulphate, filtered and concentrated in vacuo to obtain 3.86 g
of brown oil, which solidified upon standing (see Fig. S1†
1
for a H NMR spectrum).
In a typical hydrogenation experiment, the brown solid
(3.86 g) obtained as above was dissolved in 50 mL methanol
in a Schlenk flask under argon atmosphere. Palladium on
carbon (5 wt% Pd, 0.203 g) was added and the flask brought
at 1 bar hydrogen pressure by use of a gas balloon. After stir-
ring for 16 h, the Pd/C catalyst was removed by centrifugation
and filtration. The solution was concentrated in vacuo to obtain
3.82 g of brown oil, which based on the 1H NMR spectrum
(see Fig. S2†) consists exclusively of hydrogenated products.
In a typical transesterification experiment the oil as
obtained above (3.82 g) was dissolved in 50 mL methanol.
Amberlyst-15 (0.516 g) was added and the mixture refluxed
for 22 h under nitrogen atmosphere. After cooling, the
Amberlyst beads were removed by filtration and the solution
concentrated to obtain a brown oil, which based on the
1H NMR spectrum contained predominantly dimethyl adipate.
Fractional vacuum distillation (14 mbar, 87–110 °C) afforded
To further improve dimethyl adipate yields, it will be
important to reduce the amount of catechol lost to the non-
productive radical oxidation. In addition, efficient separation
and recycling of the iron catalyst, for example by hetero-
genisation, will enable a truly sustainable method for the pro-
duction of dimethyl adipate.
Acknowledgements
P.C.A.B. acknowledges the Netherlands Organization for
Scientific Research (NWO) for a Vernieuwingsimpuls Veni Grant.
Dr. Rosa Bulo is acknowledged for providing access to
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Catalysis Science & Technology
Paper
computational resources. The CatchBio program is acknowl-
edged for financial support.
25 D. D. Cox and L. Que Jr., J. Am. Chem. Soc., 1988, 110,
8085–8092.
26 W. O. Koch and H.-J. Krüger, Angew. Chem., Int. Ed. Engl.,
1995, 34, 2671–2674.
27 R. Yamahara, S. Ogo, H. Masuda and Y. Watanabe, J. Inorg.
Biochem., 2002, 88, 284–294.
Notes and references
1 F. Cavani and S. Alini, in Sustainable Industrial Chemistry,
Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim, Germany,
2009, pp. 367–425.
2 A. Castellan, J. C. J. Bart and S. Cavallaro, Catal. Today,
1991, 9, 237–254.
3 V. Hessel, I. Vural Gürsel, Q. Wang, T. Noël and J. Lang,
Chem. Eng. Technol., 2012, 35, 1184–1204.
4 A. Castellan, J. C. J. Bart and S. Cavallaro, Catal. Today,
1991, 9, 255–283.
5 R. A. Reimer, C. S. Slaten, M. Seapan, M. W. Lower and
P. E. Tomlinson, Environ. Prog., 1994, 13, 134–137.
6 S. Van de Vyver and Y. Roman-Leshkov, Catal. Sci. Technol.,
2013, 3, 1465–1479.
28 H. G. Jang, D. D. Cox and L. Que Jr., J. Am. Chem. Soc.,
1991, 113, 9200–9204.
29 R. Jastrzebski, B. M. Weckhuysen and P. C. A. Bruijnincx,
Chem. Commun., 2013, 49, 6912–6914.
30 G. J. P. Britovsek, J. England and A. J. P. White, Inorg. Chem.,
2005, 44, 8125–8134.
31 H. F. Coward and G. W. Jones, Bureau of Mines Bulletin,
1952, vol. 503, pp. 1–154.
32 B. M. Trost and P. G. McDougal, J. Org. Chem., 1984, 49,
458–468.
33 S. Seltzer and J. Hane, Bioorg. Chem., 1988, 16, 394–407.
34 M. G. Weller and U. Weser, J. Am. Chem. Soc., 1982, 104,
3752–3754.
7 K. Sato, M. Aoki and R. Noyori, Science, 1998, 281,
1646–1647.
8 M. Faber, US Pat. 4,400,468, 1983.
9 P. K. Wong, C. Li, L. P. Stubbs, M. van Meurs, D. G. Anak
Kumbang, S. C. Y. Lim and E. Drent, International Patent
Application WO 2012/134397 A1, 2012.
35 M. Pascaly, M. Duda, F. Schweppe, K. Zurlinden, F. K. Müller
and B. Krebs, J. Chem. Soc., Dalton Trans., 2001, 828–837.
36 J. A. Elvidge, R. P. Linstead, P. Sims and B. A. Orkin,
J. Chem. Soc., 1950, 2235.
37 G. E. K. Branch and M. A. Joslyn, J. Am. Chem. Soc., 1935, 57,
2388–2394.
10 K. M. Draths and J. W. Frost, J. Am. Chem. Soc., 1995, 117,
2395–2400.
38 C. R. Dawson and J. M. Nelson, J. Am. Chem. Soc., 1938, 60,
245–249.
11 J. Zakzeski, P. C. A. Bruijnincx, A. L. Jongerius and
B. M. Weckhuysen, Chem. Rev., 2010, 110, 3552–3599.
12 V. M. Roberts, V. Stein, T. Reiner, A. Lemonidou, X. Li and
J. A. Lercher, Chem. – Eur. J., 2011, 17, 5939–5948.
13 J. Zakzeski, A. L. Jongerius, P. C. A. Bruijnincx and
B. M. Weckhuysen, ChemSusChem, 2012, 5, 1602–1609.
14 Q. Song, F. Wang, J. Cai, Y. Wang, J. Zhang, W. Yu and J. Xu,
Energy Environ. Sci., 2013, 6, 994–1007.
39 N. Schweigert, A. J. B. Zehnder and R. I. L. Eggen, Environ.
Microbiol., 2001, 3, 81–91.
40 Y. Hitomi, M. Yoshida, M. Higuchi, H. Minami, T. Tanaka
and T. Funabiki, J. Inorg. Biochem., 2005, 99, 755–763.
41 M. Higuchi, Y. Hitomi, H. Minami, T. Tanaka and
T. Funabiki, Inorg. Chem., 2005, 44, 8810–8821.
42 R. Jastrzebski, M. G. Quesne, B. M. Weckhuysen,
S. P. de Visser and P. C. A. Bruijnincx, Chem. – Eur. J., 2014,
20, 15686–15691.
15 A. Baeyer, Chem. Ber., 1875, 8, 153–155.
16 W. H. Perkin, J. Chem. Soc., Trans., 1890, 57, 587–589.
17 H. D. Dakin, H. T. Clarke and E. R. Taylor, Org. Synth.,
1923, 3, 28.
18 S. B. Waghmode, G. Mahale, V. P. Patil, K. Renalson and
D. Singh, Synth. Commun., 2013, 43, 3272–3280.
19 K. Barta, G. R. Warner, E. S. Beach and P. T. Anastas, Green
Chem., 2014, 16, 191–196.
20 T. D. H. Bugg, Tetrahedron, 2003, 59, 7075–7101.
21 F. H. Vaillancourt, J. T. Bolin and L. D. Eltis, Crit. Rev.
Biochem. Mol. Biol., 2006, 41, 241–267.
22 M. Costas, M. P. Mehn, M. P. Jensen and L. Que Jr., Chem.
Rev., 2004, 104, 939–986.
43 C. B. Kretschmer, J. Nowakowska and R. Wiebe, Ind. Eng.
Chem., 1946, 38, 506–509.
44 C. F. Guerra, Theor. Chem. Acc., 1998, 99, 391–403.
45 G. te Velde, F. M. Bickelhaupt, E. J. Baerends, C. Fonseca
Guerra, S. J. A. van Gisbergen, J. G. Snijders and T. Ziegler,
J. Comput. Chem., 2001, 22, 931–967.
46 ADF2013, SCM, Theoretical Chemistry, Vrije Universiteit,
Amsterdam, The Netherlands, http://www.scm.com.
47 J. Perdew, Phys. Rev. B, 1986, 33, 8822–8824.
48 A. D. Becke, Phys. Rev. A, 1988, 38, 3098–3100.
49 E. Van Lenthe, J. Comput. Chem., 2003, 24, 1142–1156.
50 C. C. Pye and T. Ziegler, Theor. Chem. Acc., 1999, 101, 396–408.
51 S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem.
Phys., 2010, 132, 154104.
23 P. C. A. Bruijnincx, G. van Koten and R. J. M. Klein Gebbink,
Chem. Soc. Rev., 2008, 37, 2716.
24 M. W. Vetting and D. H. Ohlendorf, Structure, 2000, 8,
429–440.
52 S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem.,
2011, 32, 1456–1465.
This journal is © The Royal Society of Chemistry 2015
Catal. Sci. Technol.